In the Linux kernel, the following vulnerability has been resolved:

block: Fix potential deadlock in blk_ia_range_sysfs_show()

When being read, a sysfs attribute is already protected against removal
with the kobject node active reference counter. As a result, in
blk_ia_range_sysfs_show(), there is no need to take the queue sysfs
lock when reading the value of a range attribute. Using the queue sysfs
lock in this function creates a potential deadlock situation with the
disk removal, something that a lockdep signals with a splat when the
device is removed:

[ 760.703551] Possible unsafe locking scenario:
[ 760.703551]
[ 760.703554] CPU0 CPU1
[ 760.703556] ---- ----
[ 760.703558] lock(&q->sysfs_lock);
[ 760.703565] lock(kn->active#385);
[ 760.703573] lock(&q->sysfs_lock);
[ 760.703579] lock(kn->active#385);
[ 760.703587]
[ 760.703587] *** DEADLOCK ***

Solve this by removing the mutex_lock()/mutex_unlock() calls from
blk_ia_range_sysfs_show().

References

• https://nvd.nist.gov/vuln/detail/CVE-2022-49406
• https://git.kernel.org/stable/c/41e46b3c2aa24f755b2ae9ec4ce931ba5f0d8532
• https://git.kernel.org/stable/c/717b078bc745ba9a262abebed9806a17e8bbb77b
• https://git.kernel.org/stable/c/dc107c805cde709866b59867ef72b9390199205e